矢量中值滤波在林业TM影像处理中的应用

Application of vector median filtering in TM image processing of forestry

介绍了改进的矢量中值滤波法算法,并使用该算法对广东省九曲水林场TM图像数据进行滤波消噪处理。以图像基本统计信息、典型地类分类精度和平滑指数FI为评价指标将矢量中值滤波的处理结果与传统的滤波算法(均值滤波、中值滤波、统计滤波、自适应滤波和高斯滤波)进行了对比分析。结果表明,对林区TM图像而言,矢量中值滤波器具有出色的平滑去噪能力,并且保持了很好的分类精度,可以替代传统的滤波方法。

The vector median filtering was introduced and employed to dispose TM image data from Jiuqushui forestry region, Guangdong province. Five common filtering methods, i. e. , mean-smoothness filter, median filter, statistics filter, adaptive filter and Gauss filter, were chosen in experiment. The statistics of the characteristics of filter images,t.he classification accuracy of the typical ground classes and Filtering Index （FI） were taken as the evaluation indexes. The comparative analysis on the test results was made. Results showed that the vector median filter had a better performance on smoothing and denoising effects for TM image on forestry region. Like other common filters, it had a good accuracy of classification. This method could be used instead of the other common used filtering methods.